Temperature Trends 1990-2007: Hansen, IPCC, Obs

January 18th, 2008

Posted by: Roger Pielke, Jr.

The figure below shows linear trends in temperature for Jim Hansen’s three 1988 scenarios (in shades of blue), for the IPCC predictions issued in 1990, 1995, 2001, 2007 (in shades of green), and for four sets of observations (in shades of brown). I choose the period 1990-2007 because this is the period of overlap for all of the predictions (except IPCC 2007, which starts in 2000).

temp trends.png

Looking just at these measures of central tendency (i.e., no formal consideration of uncertainties) it seems clear that:

1. Trends in all of Hansen’s scenarios are above IPCC 1995, 2001, and 2007, as well as three of the four surface observations.

2. The outlier on surface observations, and the one consistent with Hansen’s Scenarios A and B is the NASA dataset overseen by Jim Hansen. Whatever the explanation for this, good scientific practice would have forecasting and data collection used to verify those forecasts conducted by completely separate groups.

3. Hansen’s Scenario A is very similar to IPCC 1990, which makes sense given their closeness in time, and assumptions of forcings at the time (i.e., thoughts on business-as-usual did not change much over that time).

The data for the Hansen scenarios was obtained at Climate Audit from the ongoing discussion there, and the IPCC and observational data is as described on this site over the past week or so in the forecast verification exercise that I have conducted. This is an ongoing exercise, as part of a conversation across the web, so if you have questions or comments, please share them, either here, or if our comment interface is driving you nuts (as it is with me), then comment over at Climate Audit where I’ll participate in the discussions.

8 Responses to “Temperature Trends 1990-2007: Hansen, IPCC, Obs”

  1. lucia Says:


  2. 2
  3. lucia Says:

    Is that GISS Met Station Data? Or Land Ocean?

    I’ve pretty much convinced myself that the principle of comparing like to like requires us to compare computations to Land-Ocean data whenever possible. The reason I think this is:

    1)the GCMs compute surface temperatures including areas over both the land and the ocean.
    2) We understand why surface temperatures are likely to vary more quickly than ocean values based on phenomenology.

    So, in periods of warming, measurements over land will rise faster (and this is even without considering any heat island effects.)

    Of course, as a practical matter, if we distrust the measurements over the oceans, I’d go with the land based only measurements.

  4. 3
  5. Roger Pielke, Jr. Says:

    Lucia- Thanks (and congrats for getting a comment through;-) I showed Met Stations, Land/Ocean trend is 0.22.

  6. 4
  7. Mark Bahner Says:

    Hi Roger,

    Just for clarity, can you repeat what the scenarios are for the IPCC values for 1990, 1995, 2001, and 2007?

    The scenario for 2001 and 2007 is A1F1, right?

  8. 5
  9. Roger Pielke, Jr. Says:

    Mark- Correct.

  10. 6
  11. Roger Pielke, Jr. Says:

    A comment sent in by email from Roger Cohen:

    “I think your focus on identifying metrics to track “projections” by the
    IPCC and others involved in the climate change business is a very
    worthwhile effort. Every significant business enterprise does this in
    one way or another. And since there is so much at stake here, it seems
    reasonable to insist on some kind of quantitative accountability for the
    computer-based projections from which policy recommendations have
    emerged. However, I suspect that unless the effort generates a
    sustained focus such as suggested below, it is not likely to overcome
    resistance to the concept.

    I believe the number one metric remains global average temperature
    change, flawed and incomplete that may be. As you point out, even that
    simple metric is subject to issues around the choice of a data set. For
    example, it is possible to do some statistics on the differences between
    the various data sets over a particular interval. If one compares the
    GISS data set for 2001-2007 with the UKMET (HadCRUT3) data set, one
    finds that the difference between the regression slopes over this period
    is 1.16 standard deviations, implying that there is a 97% chance that
    the UKMET set is giving a systematically lower trend than GISS over the
    past seven years. A similar analysis using the NCDC data set gives an
    86% chance that it is giving a lower systematic trend than GISS. This
    finding, combined with the problems revealed last year in the treatment
    of U.S. data and the less-than-arm’s-length relationship between Jim
    Hansen and data set preparation, reduces confidence in using the GISS
    data set in future metric analyses. And as you point out, the
    satellite-based observations are different yet. Naturally over the
    longer term, these differences will likely become unimportant, but we
    are talking here about assessing the reliability of IPCC projections
    over intermediate time frames.

    To establish a basis for agreement on how to deal with global average
    temperature and other metrics, it may be worthwhile to convene a
    workshop of parties from various disciplines who have a serious and
    sincere interest in arriving at an empirical way of assessing the
    technical merit of IPCC projections. A successful outcome of such a
    workshop and any follow up activities would be an agreement on key
    measurements and data analyses which would be followed for an indefinite
    period of time. An annual report of the state of the analyses could be
    published in a widely-read climate research journal or in an independent
    vehicle. Though universal support is not realistic, such a step could
    garner support from advocates and skeptics alike and head off much
    future cherry-picking by one side or another. Because the IPCC’s
    impartiality has been questioned in some quarters, the process of
    monitoring its own projections should not be left entirely to that
    organization. I believe that your (and that of Professor Pielke senior)
    role in keeping the system honest, while not denying human impact on
    climate, positions you to take the lead in such a project.

    Roger W. Cohen
    Durango, CO”

  12. 7
  13. Harry Haymuss Says:

    *If* global temperature would be a good way to track “global warming” the only accurate way to do it would be without any subjective corrections for, e.g., UHI.

    That means MSU. Period.

  14. 8
  15. Philips Says:

    Hansen, the IPCC and observational temperature data are being compared and analyzed here through a well read bar chart showing a linear trend in global temperature. Even a layman can gather some idea from this chart.